Spring in baseball is the time to think about pitcher-fielding practice, bunt plays and additions to your pitch mix — a moment of optimism for all 30 teams. It is also the most fruitful time of the year for contract extensions.

Over the last 15 years, more than 45 percent of contract extensions were signed in February, March or April, according to MLB Trade Rumors’ invaluable extension tracker. That’s why we’re making this Extension Week at The Athletic, looking at some of the biggest names under current team control and what an extension could look like.

Here’s how I go about this: I’ve got a big Excel spreadsheet with about 900 free-agent and extension contracts in it, as well as players’ performance in the years leading up to those contracts, as defined by FanGraphs’ wins above replacement. Whereas players get paid off home runs and saves in arbitration, I’ve found fWAR to be a solid (though not all-knowing) predictor of earnings on the open market.

I plug extension candidates into the spreadsheet, find players with similar levels of production ideally with similar amounts of team control left before they signed their deals, and work off of those comps.

Several points worth noting:

• For extensions, the team control part of this is important. If Player A looks just like Player B, except Player A has three years of arbitration ahead of him and Player B was a free agent, Player A ain’t gonna get the contract Player B did. However, what Player B made in free agency is still useful as a guide to what the later part of Player A’s contract can look like.

• For starting pitchers and position players, I look at fWAR in one-, three- and five-year samples, with a special emphasis on the shorter term. For extensions in particular, the five-year sample is less helpful because a lot of players haven’t been playing five full years.

• For relievers, I look at one- and two-year samples. What happened five years ago isn’t really relevant for such a volatile position, and I’ve found more of a recency bias in contracts here than elsewhere.

• The shortened 2020 season really is a pain in the analysis because it makes historical comparisons more difficult. After much tinkering, my method now is to take a player’s per-game fWAR from 2019 to 2021 and add 102 more games at that pace to his 2020 total. For players who hit the market just after 2020, I do the same thing using their 2018 to 2020 per-game fWAR.